A5028352910- Parallel Computing and Optimization Techniques
- Advanced Data Storage Technologies
- Distributed and Parallel Computing Systems
- Cloud Computing and Resource Management
Louisiana State University
2018-2021
Louisiana State University Agricultural Center
2020
The new challenges presented by exascale system architectures have resulted in difficulty achieving the desired scalability using traditional distributed-memory runtimes.Asynchronous many-task systems (AMT) are based on a paradigm showing promise addressing these challenges, providing application developers with productive and performant approach to programming next generation systems.HPX is C++ Library for concurrency parallelism that developed STE||AR Group, an international group of...
We study the simulation of stellar mergers, which requires complex simulations with high computational demands. have developed Octo-Tiger, a finite volume grid-based hydrodynamics code Adaptive Mesh Refinement is unique in conserving both linear and angular momentum to machine precision. To face challenge increasingly complex, diverse, heterogeneous HPC systems, Octo-Tiger relies on high-level programming abstractions. use HPX its futurization capabilities ensure scalability between nodes...
In this research, we describe the functionality of AGAS (Active Global Address Space), a subsystem HPX runtime system that is designed to handle data locality at runtime, independent hardware and architecture configuration. enables transparent global access migration, but incurs an overhead cost runtime. We present method assess performance amount impact it has on execution time Octo-Tiger application. With our assessment identify four most expensive operations in demonstrate caused by negligible.
Despite advancements in the areas of parallel and distributed computing, complexity programming on High Performance Computing (HPC) resources has deterred many domain experts, especially machine learning artificial intelligence (AI), from utilizing performance benefits such systems. Researchers scientists favor high-productivity languages to avoid inconvenience low-level costs acquiring necessary skills required for at this level. In recent years, Python, with support linear algebra...
Analyzing performance within asynchronous many-task-based runtime systems is challenging because millions of tasks are launched concurrently. Especially for long-term runs the amount data collected becomes overwhelming. We study HPX and its performance-counter framework APEX to collect energy consumption. added application-specific counters Octo-Tiger full 3D AMR astrophysics application. This enables combined visualization physical highlight bottlenecks with respect different solvers....